Symposium 1: ERP and MEG studies on phoneme discrimination,
and semantic and syntactic processing: Basic and clinical research
Symposium 2: Delayed cardiovascular recovery: A possible
alternative to reactivity models of disease
Symposium 3: The psychophysiology of emotion: The role
of stimulus relevance, motivational congruence, and coping potential
Panel Discussion 1: Entering the domain of the dense
electrode array: Conceptual and methdological problems
Symposium 4: Ambulatory psychophysiological assessment:
Methodological challenges and clinical opportunities
Symposium 5: Psychophysiology in the clinic: The anxiety
disorders
Symposium 6: Ten years after: Paul Obrist's legacy
to cardiovascular psychophysiology
Symposium 7: ERPs to subliminal stimuli: Markers for
unconscious processes
Panel Discussion 2: Review and funding of psychophysiological
research: Perspectives, opportunities, and strategies
Symposium 8: Psychophysiological studies of biobehavioral
mediators of risk for alcohol abuse
Symposium 9: Synthetic Emotion
Symposium 10: Steady-state evoked potentials: New
methodological approaches and applications
Symposium 11: The role of cortisol in memory, attention,
and emotion
Special Panel Discussion: Psychophysiology: The Shape
of its Future
ERP and MEG studies on phoneme discrimination, and semantic and
syntactic processing: Basic and clinical research
Chairs: John F. Connolly & Kimmo Alho
Participants: Kimmo Alho, John Connolly, Myong Yoon, Ryan D'Arcy
The intent of this symposium is to present new findings of ERP and MEG studies
in language comprehension in basic research contexts and in clinical applications.
The goal is to present studies that move along the continuum of complexity
of processing beginning with basic pre-attentive processing of speech components,
moving to phonological and semantic processing of simple sentences, then
to neural mechanisms involved in the processing of compound sentences that
have semantic and/or syntactic violations in the first and/or second sentence
segment, and finishing with a presentation of the clinical applications
of this basic research. The first presentation deals with ERP/MEG findings
demonstrating MMN responses to phonemes that reflect pre-attentive processes
that are lateralized to the left hemisphere. The second talk will briefly
review the phonological mismatch negativity (PMN) and present recent findings
demonstrating its temporal sensitivity to the "uniqueness point"
and work conducted in Helsinki (Connolly, Alho et al., in preparation) demonstrating
its lateralization to the left temporal cortex. The third presentation reports
recent findings using compound sentences to study the manner in which semantic
and syntactic processes interact in the comprehension of text and the importance
of sentence context on these effects. The final paper describes recent advances
in the application of knowledge derived from the basic research exemplified
in the first three papers to the investigation of clinical syndromes characterized
by language dysfunction or to the assessment of intellectual functioning
in non-communicative individuals.
Mismatch negativity reveals hemispheric lateralization of preattentive
speech processing: Recent electrical and magnetic recordings
Kimmo Alho
University of Helsinki
The mismatch negativity (MMN) is an event-related brain potential (ERP)
elicited by deviant sounds occurring in a sequence of repeating sounds.
Recent studies from our research unit indicate hemispheric lateralization
in processing of unattended speech sounds reflected by MMN and its magnetoencephalographic
(MEG) counterpart (MMNm). Naatanen et al. (Nature, 1997, 385, 432-434) found
that MMNm is generated with a larger amplitude in the left than in the right
auditory cortex by deviant vowels when they are prototypical phonemes of
the subject's native language. Tervaniemi et al. (in preparation) observed
a stronger MMNm in the right than in the left auditory cortex to a change
in a musical chord whereas the opposite was true for a vowel change. Alho
et al. (in preparation) found a left-hemisphere dominant MMNm also for changes
in a consonant- vowel syllable. Moreover, source activity modelled by Rinne
et al. (in preparation) for electrical MMNs to changes in eight types of
sounds ranging from sinusoidal to phonetic indicated a left-hemisphere dominance
for MMNs in the phonetic end of the stimulus continuum and a right-hemisphere
dominance in the other end of the continuum. While positron emission tomography
(PET) and functional magnetic resonance imaging (fMRI) have shown a left-hemisphere
dominance for attentive processing of speech sounds, the aforementioned
MMN studies indicate hemispheric lateralization already at the pre-attentive
level of speech processing.
Sensitivity of the phonological mismatch negativity to phonemic violations
and its lateralization to the left temporal cortex: ERP and MEG findings.
John F. Connolly
Dalhousie University
The phonological mismatch negativity (PMN) is an event-related potential
component that has been shown in several studies (e.g., Connolly & Phillips,
J. Cognitive Neuroscience, 1994, 6, 256-266; Connolly et al., J. Clinical
& Experimental Neuropsychology, 1995, 17, 548-565) to be sensitive to the
initial phoneme of a word that ends a spoken sentence incongruously or congruously
but unexpectedly. The PMN has been temporally and functionally separated
from the N400 and has been shown to be specific to the auditory modality
(Connolly et al., Electroencephalography & Clinical Neurophysiology, 1995,
94, 276-287). More recently, the additivity of these two components was
evaluated (Kutas & Hillyard, Biological Psychology, 1980, 11, 99-116); results
suggested the two components are additive and lend further support to the
independence of the PMN and N400. Also, recent work is described that manipulated
the point at which sentence-ending words became contextually incongruous;
unexpected/incorrect phonemes occurring at word onset (The piano was out
of leaves.), at word midpoint (The gambler had a bad streak of luggage.)
or at a late position in the word (He mailed the letter without a stampede.).
This manipulation in point of phoneme deviance resulted in a progressive
delay in the occurrence of these components (Connolly & Randhawa, in preparation).
Finally, recent work conducted with colleagues in Helsinki (Connolly, Alho
et al., in preparation) using both electrical (ERP) and magnetoencephalographic
(MEG) techniques has replicated the electrical results from earlier work
and localized the PMN to the left temporal cortex. In sum, these findings
support both ERP and MEG techniques in the identification of the functional
components of speech comprehension as well as their neuroanatomical localization.
Distinctive neural responses to semantic anomaly and/or inflectional
number-disagreement in compound sentences: An ERP study
Myong G. Yoon & John F. Connolly
Dalhousie University
We studied neural responses to semantic anomaly and/or inflectional number
disagreement in compound sentences during reading by normal subjects. [1]
The ERPs of semantically anomalous (but inflectionally correct) verbs in
the final clauses of the compound sentences showed negative shifts (N400)
between 350 and 600 ms (peak; 450 ms) with respect to those for appropriate
correct verbs. When the semantically anomalous verbs appeared in the first
clauses, their ERPs showed the negative shifts (N400) about 100 ms earlier
between 250 and 450 ms (peak; 350 ms). [2] The ERPs of inflectionally incorrect
(but semantically appropriate) verbs in the first clauses showed positive
shifts (peak; 600 ms: P600) with respect to those for correct verbs. [3]
The ERPs of semantically anomalous and inflectionally incorrect verbs were
either P600 type or N400 type, depending on the locations of recording electrodes.
[4] The ERPs of semantically anomalous and inflectionally incorrect verbs
showed large positive shifts (P600) compared with the ERPs of semantically
congruous and inflectionally incorrect verbs. This result reveals that neural
responses to the inflectional disagreement are greatly enhanced by the semantic
anomaly of the verbs. [5] The ERPs of the object words (located at the ends
of sentences) of semantically anomalous verbs showed very large negative
shifts during a period between 300 and 750 ms. These results reveal the
prominent influences of various sentential contexts on the human brain's
neural responses to linguistic stimuli.
The application of event-related potentials (ERPs) in the assessment
of receptive language functions
Ryan C.N. D'Arcy, John F. Connolly, Joseph M. Byrne, & Christopher A. Dywan
Dalhousie University & Izaak Walton Killam Hospital
While there is a long and distinguished history of basic research with event-related
potentials (ERPs), more concentration on the clinical applications of this
technique is paramount. Language is well suited for investigations of an
applied nature because various ERP components associated with linguistic
processing have been identified. Further, receptive language dysfunction
is frequently noted in aphasic individuals suffering from stroke, but due
to communicative limitations, may not be assessable with traditional psychometric
measures. Accordingly, we have explored the feasibility of adapting standardized
neuropsychological measures of language for computer presentation with ERPs.
These investigations have demonstrated that components related to speech
perception and recognition memory (e.g., N400 and P300) can be elicited
by standardized test stimuli and thus utilized to evaluate language independent
of behavior. Specifically, studies of vocabulary and comprehension have
demonstrated that: 1) test items within an individual's ability, but not
beyond, can be differentiated on the basis of their neural responses; 2)
subtle increases in difficulty do not influence behavioral responses, but
reliably attenuate neural responses; and 3) these results exist at the individual
level, thus maintaining clinical applicability. The potential implications
of this novel technique will be elucidated and some recent experimental
findings with the Token Test (De Renzi & Vignolo, 1962, Brain, 85, 665-678)
and the Psycholinguistic Assessment of Language Processing in Aphasia (Kay
et al., 1992) will be presented. The objective of this work concerns the
development of an electrophysiological aphasia battery and the extension
of this approach into other realms of cognition (e.g., memory/amnesia).
Delayed cardiovascular recovery: A possible alternative to reactivity
models of disease
Chair: William Gerin
Participants: David Sheffield, Laura Glynn, Julian Thayer, Paul Mills
Discussant: Nicholas Christenfeld
Cardiovascular reactivity remains the predominant psychophysiological model
of stress effects on cardiovascular disease. However, several studies now
suggest that cardiovascular recovery of pre-stress resting levels may provide
important prognostic information, independent of that provided by reactivity-testing.
As with reactivity, it is important to understand the determinants of delayed
recovery at several levels, as well as to examine the effects of delayed
recovery prospectively. This panel covers a broad range of issues related
to recovery. Paul Mills describes the role of adrenergic receptors in regulating
blood pressure recovery, and Julian Thayer reports on a model of anger inhibition
and cardiovascular recovery that may shed light on mechanisms underlying
blood pressure regulation, with a focus on vagal influences. Both authors
report on Black-White differences which may help us to understand the greater
prevalence of hypertension in Black Americans, as well as the role of individual
differences in anger expression (Thayer) that may contribute to delayed
recovery. Laura Glynn describes the role of situational factors, specifically
distraction, that may prolong blood pressure elevations in response to stress;
and David Sheffield reports on a prospective study in which cardiovascular
recovery accounted for a greater proportion of variance than did reactivity
in blood pressure 5 years later, thus providing validation for the clinical
importance of measuring recovery. Nicko Christenfeld will discuss the papers,
with reference to methodological and measurement issues, and will provide
an integration of the issues raised by the presenters.
Does blood pressure recovery from an active mental stressor predict future
blood pressure? A comparison with blood pressure reactivity
David Sheffield1, George Davey Smith2, Douglas Carroll3, Martin J. Shipley4,
& Michael G. Marmot4
1University of North Carolina, 2University of Bristol, 3University of Birmingham,
4University College London
The evidence linking the magnitude of cardiovascular reactions to stress
and future blood pressure status is mixed. Few studies, though, have examined
recovery in this context. We report on a prospective study which allows
us to compare the contributions of reactivity and recovery to the prediction
of future blood pressure. A sample of 969 British male public servants (average
age 44.0 years), revealed as normotensive at an initial screening of blood
pressure, had their blood pressure monitored in the laboratory at rest,
while undertaking an active mental stress task, Raven's matrices, and after
the task. Blood pressure reactivity was calculated as laboratory task-rest
difference; recovery was defined as level post-task. Follow-up blood pressure
screening was conducted, on average, five years later. Analysis was by step-wise
multiple regression, with age always entered first. In the case of follow-up
systolic blood pressure, a model including age, initial screening blood
pressure and resting laboratory blood pressure accounted for 42 per cent
of the variance in follow-up screening blood pressure, and, while systolic
blood pressure reactions entered the equation, they accounted for less than
one per cent of the variance; recovery did not enter the equation. In the
case of follow-up diastolic pressure, age and diastolic blood pressure at
initial screening accounted for 35 per cent of the variance. In this case,
though, diastolic blood pressure recovery entered the equation, accounting
for an additional three per cent of the variance. These analyses suggest
that recovery may have greater prognostic significance than reactivity.
Emotion, rumination and the rate of blood pressure recovery from stress
Laura Glynn1, Nicholas Christenfeld1, & William Gerin2
1University of California-San Diego, 2Cornell University Medical Center-The
New York Hospital
While many emotional stressors are themselves brief--being cut off on the
freeway, or insulted by a stranger--their effects can be long-lived. One
factor which may influence this duration is rumination. To examine this
issue, we conducted two experiments on BP recovery from laboratory stressors.
In the first experiment, we hypothesized that emotional stresssors may lend
themselves more readily to rumination than non-emotional, physical stressors,
and so produce slower cardiovascular recovery. Seventy-two subjects experienced
one of four stressors. Blood pressure was continuously monitored during
baseline, stress, and recovery periods. The reactivity produced by the emotional
tasks (mental arithmetic with harassment and shock threat) was the same
as that produced by the physical tasks (walking in place and cold pressor).
However, at the end of the ten minute recovery period, the physical task
subjects were within 5 mmHg SBP of baseline, but the emotional task subjects
were still elevated 15 mmHg. In the second experiment, we manipulated rumination
directly. All subjects experienced mental arithmetic with harassment. Half
were given a distractor task (a questionnaire) designed to prevent rumination,
and half were simply left alone. Those with the distraction were within
5 mmHg two minutes after the stressor, while those left to ruminate were,
ten minutes later, still 15 mmHg above baseline. These data suggest that
rumination can play a major role in recovery to baseline levels. If CVR
is involved in later CVD, it seems likely that the effect of stressors can
be magnified considerably by processes such as rumination.
Anger expression, ethnicity, and cardiovascular recovery
Nancy Dorr1, Julian F. Thayer1, & Jos F. Brosschot2
1University of Missouri-Columbia, 2University of Amsterdam
Recently, Brosschot and Thayer (1997) proposed a model of anger inhibition
and cardiovascular recovery that suggested that differences in socially
accepted forms of anger expression may help to account for the large black-white
difference in hypertension. Specifically, previous research has shown that
socially dominant individuals recovered more quickly from an anger instigation
when they had the opportunity to respond with expressions of anger. Conversely,
socially non-dominant individuals recovered more quickly when they responded
with anger inhibition. We tested this model in the context of responses
to racist and non-racist provocation in a group of college undergraduates.
Forty-six male participants (23 white, 23 black) engaged in two debates
with a confederate. One debate had a racist theme and one debate had a non-racist
theme. In both cases, the participant was provoked by the confederate. Systolic
and diastolic blood pressures were recorded during the debate and during
a recovery period which followed each debate. Anger expression was manipulated
by having half of the participants rate the confederate (anger expression)
and half of the participants rate their best friend (anger inhibition) preceding
the recovery period. Results indicated that white participants recovered
more quickly when they had an opportunity to express their anger whereas
blacks recovery more quickly when they inhibited their anger, (SBP: F(1,37)
= 4.66, p<.05; DBP: F(1,37)= 4.96, p<.05). These results are consistent
with the notion that socially acceptable forms of anger expression may contribute
to individual differences in cardiovascular recovery. Underlying mechanisms
including the role of the vagus in BP regulation will be discussed.
Adrenergic mechanisms underlying blood pressure (BP) reactivity and recovery
Paul J. Mills, Michael G. Ziegler, & Joel E. Dimsdale
University of California-San Diego
Adrenergic physiology plays an important role in regulating cardiovascular
responses to and recovery from stress. Data from two separate studies are
presented examining these phenomena. In the first study, interindividual
variations in adrenergic receptors and agonists were examined for their
association with BP responses to an acute laboratory stressor. In the second
study, BP recovery and adrenergic receptors were examined following a 2-3
day stay at a Clinical Research Center. Hospitalization routinely results
in a lowering (recovery) of chronically high BP and this study examined
potential adrenergic mechanisms underlying this phenomenon. In study 1,
50 hypertensive and normotensive Black and White individuals had their beta-
and alpha-adrenergic receptor sensitivity determined through agonist infusions.
Norepinephrine, epinephrine, BP and heart rate responses to a standardized
mental arithmetic task were then obtained. Distinct subgroups of individuals
were identified, with BP and heart rate responses determined by the sensitivity
of adrenergic receptors and the accompanying catecholamine responses. Relatively
insensitive receptors generated sizeable end-organ responses in the presence
of high catecholamine concentrations and, conversely, more sensitive receptors
generated sizeable end-organ responses in the presence of more moderate
catecholamine responses. In study 2, 88 Black and 77 White hypertensive
and normotensive individuals were studied following a 2-3 day hospitalization.
Blacks (especially Black hypertensives), as compared to Whites, showed a
smaller decline in both systolic (p<0.01) and diastolic (p=0.05) BP following
hospitalization. Black hypertensives also showed increased beta-adrenergic
receptor sensitivity (p=0.02). Together, the combined data provide evidence
for an adrenergic role in acute BP responsivity and recovery.
The psychophysiology of emotion: The role of stimulus relevance,
motivational congruence, and coping potential
Chairs: Arvid Kappas & Craig A. Smith
Participants: Craig A. Smith, Joe Tomaka, Rex Wright, Arvid Kappas
Discussant: Gregory J. McHugo
Recently, interest in affective processes has burgeoned, stimulated by findings
concerning the influence of positive and negative mood states on the startle
reflex. Yet, some feel that reducing affect to a single valence dimension
is too reductionistic because distinct emotional states of the same valence
clearly differ with regard to expressive and subjective responses, and,
according to numerous hypotheses, physiological activity. In the years since
SPR featured a panel discussion on "theoretical issues in the psychophysiology
of emotion" (1988) theoretical developments have been made, many in
the context of appraisal theories of emotion. These theories postulate that
physiological responses are a function of the importance and the motivational
congruence of a stimulus for an individual as well as his/her perceived
coping potential, and that these responses are organized around the implications
of these appraisals for adaptive action. This perspective accords well with
notions advanced by others, such as Obrist and Brehm. Because, it holds
physiological activity in emotion to be a function of the individual's evaluation
of the affective stimulus, this approach may be particularly useful in interpreting
"troublesome" observations regarding physiological differentiation
of emotions, such as an apparent lack of stimulus specificity or large individual
differences in responding. The symposium presents different theoretical
and empirical approaches that consider the role of stimulus relevance, motivational
congruence, and coping potential on physiological responses in an attempt
to go beyond positive and negative mood states as discriminators of emotion,
and to begin to delineate the organization of physiological activity in
emotion.
Appraisal as an organizing principle
Craig A. Smith
Vanderbilt University
An important proposition that has emerged from appraisal theory is that
the physiological activity associated with various emotions is organized
around the adaptational implications of the appraisals that give rise to
the emotions. On this view, emotions function to motivate individuals to
respond to adaptationally relevant circumstances that have been identified
through appraisal, and the physiological activity in emotion serves to support
physically these coping responses. Although these general ideas have been
proposed for at least thirty years (e.g., Lazarus, 1968, Nebraska Symposium
on Motivation, 175-266), it is only within the past decade that the specific
implications of these propositions have begun to be developed and empirically
tested (e.g., Smith, 1989, Journal of Personality and Social Psychology,
339-353). Even today, progress in testing this approach is slow, perhaps
because the theory and its implications for physiological activity in emotion
are not widely known.In this presentation, I first consider the fundamental
assumptions and principles of appraisal theory, as well as the theoretical
reasons for expecting emotion-related physiological activity to be linked
to appraisal. Then, the derivation of specific, testable hypotheses from
the more general theoretical principles is considered, and the nature of
such hypotheses is contrasted with a more traditional categorical approach
of testing for general emotion-specific patterns of physiological activity.
Finally, the promise of the appraisal approach to understanding the organization
of physiological activity in emotion is illustrated with recent findings
that explicitly link appraisal information to the dynamics of spontaneous
autonomic activity (e.g., finger temperature, skin conductance) during ongoing
problem-solving tasks.
Relation of threat and challenge to emotion and coping theory
Joe Tomaka, Julie A. Penley, & Rebecca Palacios
University of Texas at El Paso
Threat and challenge are distinct patterns of stress-related responding
that result from cognitive appraisals of situational demands considered
in relation to appraisals of ability to cope with such demands (see Tomaka,
Blascovich, Kelsey, & Leitten, 1993, Journal of Personality and Social Psychology,
248-260). In addition, we have demonstrated that threat and challenge have
distinct affective, behavioral, and cardiovascular consequences. For example,
threat appraisals result in negative affect, poor task performance, and
a pattern of cardiovascular response that includes moderately increased
cardiac output coupled with an increase in systemic vascular resistance.
Challenge appraisals, in contrast, result in less negative affect, better
task performance, and a pattern of cardiovascular response that includes
increased cardiac output, coupled with a decline in systemic vascular resistance.
Research has shown threat and challenge response patterns can be produced
experimentally and that such patterns are reliably related to aspects of
personality.In this presentation we briefly discuss the nature of threat
and challenge responses and then turn to more recent research examining
the motivational underpinnings of such responses and research relating threat
and challenge to the broader context of emotion theory. We conclude with
suggestions for applying threat and challenge research to other areas of
psychophysiological inquiry. Throughout this presentation, we emphasize
the advantage of using patterns of autonomic responses over single parameter
investigations and emphasize the use of multiple converging measures--including
physiological and nonphysiological--to identify psychophysiological constructs.
Perceived performance capacity and cardiovascular response to a fixed
behavioral challenge
Rex A. Wright
University of Alabama at Birmingham
An analysis is presented that suggests that effort and associated cardiovascular
(CV) responses to a fixed (specific) behavioral challenge are determined
jointly by the difficulty of the challenge and the performer's appraised
ability in the relevant performance realm. As difficulty increases, effort
and CV responsivity should first be greater for low ability - than high
ability individuals, then be greater for high ability- than low ability
individuals, then low and equivalent for both groups. The exact points at
which the perceived-ability/CV-response relation reverses and then disappears
should depend on the importance of success. Where success importance is
relatively low, the changes should occur at relatively low difficulty levels;
where success importance is relatively high, the changes should occur at
relatively high difficulty levels. The analysis is supported by data from
a variety of recent experiments, and would seem to have implications for
a number psychophysiological phenomena, including sex and race differences
in CV response.
His master's voice: Acoustic analysis of spontaneous vocalizations in
an ongoing active coping task
Arvid Kappas
Laval University
There is still little evidence supporting the hypothesized relationship
between expressive behavior and outcomes of cognitive evaluations along
appraisal dimensions. For example, Scherer (1986, Psychological Bulletin,
143-165) hypothesized that goal obstruction is accompanied by an increase
in the higher frequency spectrum of the voice, in the fundamental frequency
(F0), as well as in period-to-period perturbations of F0 and amplitude (jitter
and shimmer). However, these relations have not yet been tested systematically.
One of the difficulties in studying such hypotheses is to find a paradigm
that is well suited to assessing facial and vocal activity in an ongoing
active coping task. I have developed a video-game paradigm to test the impact
of manipulating goal congruence and coping potential on facial and vocal
activity. Specifically, participants controlled a Pacman-type video game
using a limited set of voice commands "gauche", "droite",
"haut", "bas", while multiple physiological measures
were taken (SC activity, IBI, finger temperature, facial EMG at the Corrugator
Supercilii site). Recording highly standardized vocalizations permits averaging
within words, accounting for the variance in acoustic changes that are due
to their phonetic structure. Because the voice commands were the direct
mode of interacting with the game, I consider these vocalizations as spontaneous
and adequate for the test of the hypothetical relations between appraisals
and acoustical changes. Specifically, I will present data relating appraisals
of goal congruence and coping potential to event-related physiological activity
including vocal changes, ANS responses, and EMG activity at the Corrugator
Supercilii site. The advantages and disadvantages of using acoustic analyses
of spontaneous vocalizations to study affective processes will be discussed.
Entering the domain of the dense electrode array: Conceptual
and methdological problems
Chair: Emanuel Donchin
Participants: Emanuel Donchin, Don M. Tucker, Kevin M. Spencer, Thomas R.
Elbert
Discussant: Greg McCarthy
An increasing number of laboratories are beginning to use "dense electrode
arrays" for recording EEG. Such arrays cover the scalp with 64 or,
more recently, with 128 electrodes. These recording arrays are used because
they are likely to allow a detailed examination of the componential structure
of the ERPs. The dense arrays can also be used to support numerical methods
for source localization such as BESA. The realization of these possibilities
requires, however, that we solve the serious problems presented by the massive
data base that is acquired when such arrays are recorded. The participants
in this panel will review the special challenges presented by the dense
arrays. Some of the pitfalls encountered when working with dense arrays
will be reviewed, and guidelines for avoiding these pitfalls will be discussed.
The challenges presented by dense electrode arrays
Emanuel Donchin
University of Illinois at Urbana-Champaign
This presentation will introduce the panel. I will review the history of
the use of multiple electrodes in ERP research, the problems that arose
when scalp distribution became a defining attribute of ERP components, and
the solutions that have been adopted by different investigators. The manner
in which these issues are transformed when one uses 128 electrodes will
be examined and the special issues that arise in this context will be reviewed.
In particular I will review critically attempts to adopt to the mapping
of ERPs methods that derive from fMRI research. The visual inspection of
selected maps, or of movies, will also receive a critical examination. It
will be my contention that, as is true for data obtained with a single electrode,
there is no substitute for a detailed quantitative analysis of the data
obtained with dense arrays. Criteria for evaluating approaches to the quantification
of brain maps will be proposed.
Spatial Nyquist of the human EEG
Don M. Tucker, Ramesh Srinivasan, & Michael Murias
University of Oregon and Electrical Geodesics, Inc.
Measuring brain electrical activity at the scalp surface requires adequate
sampling of the electrical fields in their spatial as well as temporal extent.
Temporal sampling (A/D conversion) is defined by the Nyquist theorem: The
sampling rate must be more than twice the highest frequency of the signal.
To be accurate, spatial sampling must also follow the Nyquist theorem, with
an electrode density that is more than twice the spatial period (frequency)
of the brain's electrical fields. In simulation studies with a 4-sphere
model (estimating the conductivities of brain, CSF, skull, and scalp), we
represented field patterns with differing spatial frequencies by sets of
spherical harmonics. These studies showed that the highest spatial period
that can be resolved with a 32-channel montage (distributed across the surface
of an average-sized head) is 14 cm, defining a topographic feature of 7
cm, about the size of a lobe of the brain. If there is greater spatial detail
than this in the scalp EEG, it cannot be measured by a 32-channel recording,
and it will be aliased into the lower spatial frequencies. In empirical
studies with 128-channel EEG recordings, we sub-sampled visual ERPs, motor
ERPs, and various EEG phenomena to examine the adequacy of 64-, 32-, and
19-channel montages. These data showed substantial detail in individual
subjects' scalp topographies that was often distorted in interpolations
from 64-channel subsamples, and that was severely aliased by 32- and 19-channel
montages.
Temporal-spatial analysis of the late positive components of the ERP
Kevin Spencer, Joseph Dien, & Emanuel Donchin
University of Illinois at Urbana-Champaign
Several late positive components of the event-related brain potential (ERP)
have been reported as responses to deviant stimuli: the P300, Slow Wave,
P3a, and novelty P3. The relationships between these components were investigated
using a high-density electrode array for EEG recording. Temporal and spatial
principal components analyses (PCAs) were used to reduce the dimensionality
of the data (15 subjects X 129 electrodes X 252 timepoints X 9 stimuli),
and to estimate component activity.15 subjects were studied in 4 experimental
conditions. In two "passive" conditions, subjects solved a word
puzzle or read a book while being presented with an irrelevant auditory
oddball series. In two "active" conditions, subjects responded
to rare target stimuli in two auditory oddball series containing rare and
frequent stimuli, or rare, frequent, and "novel" rare stimuli.Temporal
PCA produced temporal factors for the Slow Wave and for the other three
late positive components combined. The spatial patterns of the P300, P3a,
and novelty P3 were then dissociated by spatial PCA, which produced two
main spatial factors with anterior and posterior topographies that differed
in their relative contributions to the components. We conclude that the
novel stimuli elicit both a P300 component and a primarily frontal "novelty
P3" component. Our data suggest that the "P3a" and the "novelty
P3" are distinct ERP components.
Mapping EEG-potentials: A strategy for uncovering cortical sources
Markus Junghoefer, Thomas Elbert, & Brigitte Rockstroh
University of Konstanz
For any given distribution of the electrical potential on the surface of
the head, there exists an infinite number of possible source configurations.
This law presents a major problem when determining source locations. A second
theorem, however, maintains that the potential distribution on any closed
surface that encloses all the generators can be determined uniquely from
the surface potential map. Therefore, the blurring of the electrical potential
distribution which is caused by the high conductivity of the skull can in
principle be compensated. An alternative to this "cortical mapping"
which also compensates the spatial low pass that is characteristic of the
transformation between cortex and scalp is current source density (CSD)
calculation.In this presentation, we describe a uniform method for calculating
the interpolation of scalp EEG potential distribution, the CSD, the potential
distribution on the surface of the brain, and its CSD. It will be shown
that interpolation and deblurring methods such as CSD or cortical mapping
are not independent of the inverse. Not only the resolution but also the
accuracy of these techniques, especially those of deblurring, depend greatly
on the spatial sampling rate (i.e., the number of electrodes). Using examples
from simulated and real (64 and 128 channels) data it can be shown that
the application of more than 100 EEG channels is necessary to guarantee
a reasonable accuracy in the calculations of CSD or cortical mapping. Likewise,
it can be shown that using more than 250 electrodes does not improve the
resolution.
Ambulatory psychophysiological assessment: Methodological challenges
and clinical opportunities
Chairs: Thomas W. Kamarck & Kevin T. Larkin
Participants: Thomas W. Kamarck, Jochen Fahrenberg, David Sheps, Kevin T.
Larkin
Discussant: Derek Johnston
Research over the past decade suggests that ambulatory assessment may provide
independent information about psychophysiological functioning not available
from laboratory data alone. Ambulatory monitoring studies present a number
of methodological challenges; challenges which are increasingly surmountable
with the advent of recent technological and methodological advances. At
the same time, these studies provide important opportunities for enhancing
our understanding of clinical disorders such as hypertension and cardiovascular
disease. The presentations will focus on recent developments in ambulatory
psychophysiological assessment, highlighting both the challenges and the
opportunities associated with these methods. Dr. Kamarck will present findings
from his research employing a computerized strategy for simultaneously gathering
diary ratings of emotional activation and cardiovascular measures. In Dr.
Fahrenberg's presentation, empirical evidence outlining the use of accelerometry
devices for measuring posture and movement during ambulatory monitoring
periods will be discussed. Using coronary artery disease patients, Dr. Sheps
will present findings using ambulatory monitoring of cardiovascular ischemia
to examine how depressive mood is related to ischemic events occurring during
daily living. The correspondence between clinic and ambulatory measures
of blood pressure in defining essential hypertension will be presented by
Dr. Larkin. Finally, Dr. Johnston will serve as discussant and integrate
major findings from the papers presented with a focus upon suggestions for
future empirical work.
The effects of psychosocial influences on ambulatory blood pressure:
Contrasting different measurement and data analytic strategies
Thomas W. Kamarck, Saul S. Shiffman, Leslie Smithline, Hayley Thompson,
Jeff Goodie, Jean Paty, Mary Ann Gnys, & Jon Kassel
University of Pittsburgh
Measures of ambulatory blood pressure (ABP) may provide independent markers
of cardiovascular risk; previous reports show that psychosocial factors
(mood, work stress, and social support), in turn, may exert important influences
on ABP. Previous studies have used a variety of approaches to measure these
influences. Here we examine the consequences of different strategies for
assessing ABP determinants, using data from a sample of healthy community
adults. 120 subjects (54 % female, 50 % African American) were assessed
at 45 minute intervals for 6 days using an automated ABP device and a palmtop
computer for recording concurrent behavioral states, yielding an average
of 109 observations for each participant. Our data suggest that: (a) individual
differences in average negative affect and arousal (between-subject assessments)
are not associated with ABP differences, whereas fluctuations in emotional
activation (within-subject assessments) show substantial associations with
cardiovascular activity in the natural environment, (b) a number of time-varying
covariates (e.g., activity, posture) alter but do not eliminate the effects
of emotional activation on ABP, and c) there are significant individual
differences in the effects of emotional activation on ambulatory cardiovascular
activity. These differences were associated, in a subsample, with laboratory
measures of cardiovascular reactivity taken on the same individuals. These
data demonstrate the importance of considering temporal, situational, and
person factors in assessing the psychosocial determinants of ABP, and they
illustrate some of the merits of a computerized approach for ambulatory
diary assessment.
Posture, motions and tremor: Detection by calibrated accelerometry, DC-/AC-splitting,
and joint time-frequency analysis
Jochen Fahrenberg, Friedrich Foerster, Manfred Smeja, & Wolfgang Mueller
University of Freiburg
Assessment of posture and motion is an essential issue in ambulatory monitoring
since physiological responses, for example changes in heart rate or blood
pressure, may be due to changes in physical activity and posture. Continuous
recordings (24-hour) of posture and motion can be generally useful in behavior
assessment. Calibrated piezoresistive accelerometer devices were employed
in these investigations. The slow motion (DC-) component allows for assessment
of change in position referring to the gravitational axis; the AC-component
represents acceleration along the sensitive axis of the device. Recordings
were made in 24 subjects for nine conditions: sitting, standing, lying supine,
sitting and talking, sitting and typing on a PC keyboard, walking, climbing
stairs, walking down stairs, and cycling (all repeated once). The recordings
were continued outside the laboratory. A participant observer recorded duration
and type of activity of subjects. The Vitaport2 recorder/analyzer system
(Vitaport EDV Systeme GmbH) was used (placement of sensors: sternum, dorsum
of lower arm segment, frontal aspect of thigh and lower leg, respectively).
The findings indicated that the classification of posture and motion based
upon the accelerometer recordings was highly reliable. The agreement between
behavior observation and kinematic analysis was satisfactory, although discrepancies
existed in some subjects. This methodology was also employed to assess tremor
activity in 22 patients with Parkinson's Disease during laboratory tasks
and 24-hour monitoring using joint time-frequency analysis. A short account
on findings regarding frequency and amplitude of tremor, and distribution
of tremor episodes across 24 hour recordings will be presented.
Depressed mood is related to heart rate variability and ischemia during
daily life
David S. Sheps & David Sheffield
University of North Carolina at Chapel Hill
Depression and depressed mood have been associated with subsequent morbidity
and mortality in patients with coronary artery disease (CAD). One explanation
for these findings is that patients with depressed mood might display exaggerated
sympathetic activity resulting in more myocardial ischemia during daily
life. To examine this possibility, 81 patients with documented CAD (61 men,
20 women; mean age = 61.5 years) underwent 24-48 hour ambulatory ECG monitoring
during normal daily activities. Depressed mood was assessed by the depression
scale of the Minnesota Multiphasic Personality Inventory. An ischemic episode
was defined as at least 1 mm horizontal or downsloping ST segment depression
at 0.08 seconds after the J-point that lasted at least 1 minute. Depressed
mood was related to the number of ischemic episodes/24 hours (r=.23, p<.05)
and modestly related to the total duration of ischemic episodes/24 hours
(r=.19, p<.10). These relationships remained largely unaltered after
adjustment for disease severity (number of occluded vessels) and medications.
Heart rate variability was determined in a subsample of 35 patients. Depressed
mood was related to average heart rate (r=.52, p<.01) and inversely related
to SDNN (r=-.53, p<.01). Again, adjusting for disease severity or medications
did not alter the strength of these relationships. Finally, average heart
rate was modestly related to the number (r=.34, p<.10) and duration (r=.32,
p<.10) of ischemic episodes/24 hours; SDNN was unrelated. These analyses
suggest that depressed mood may be related to increased sympathetic tone
and more myocardial ischemia during daily life and may account for the observed
relationship between depression and mortality.
The "white coat" effect: False positives and false negatives
in the assessment and diagnosis of essential hypertension
Kevin T. Larkin, Scott L. Schauss, & D. Michael Elnicki
West Virginia University
Blood pressure (BP) determinations in the clinic often bear little resemblance
to BP measured through ambulatory or home monitoring. Although a great deal
of attention has been paid to 'white coat' hypertensive (WCH) patients (persons
who have high BPs in the clinic but normal arterial pressures at home or
work), little is known about patients whose BP is within normal limits in
the clinic, but whose average daily arterial pressures fall within the hypertensive
range, i.e., 'white coat' normotensives (WCN). The purpose of this study
was to investigate the frequency of both false positive (WCH) and false
negative (WCN) determinations of hypertensive status using clinic BPs with
ambulatory BPs as a reference. Sixty-six community volunteers were recruited
to participate in a BP assessment study. Auscultatory BPs were obtained
in the clinic, home BPs were measured over a one-week period, and ambulatory
BPs were obtained over a 24-hour period. Although clinic, home, and ambulatory
measures of BP were all highly correlated, categorization using JNC V blood
pressure guidelines was inconsistent. Nine volunteers were categorized as
WCH and 12 were categorized as WCN using mean clinic and ambulatory BPs.
WCN patients were more mentally active and slept less than WCH patients,
and were more physically active than normotensive volunteers during ambulatory
monitoring. Similar findings were observed with analysis of home BPs. This
study demonstrated that establishing a diagnosis of hypertension based solely
upon clinic BP measures results in a significant number of both false positive
and false negative diagnoses using ambulatory or home BPs as standard referents.
Psychophysiology in the clinic: The anxiety disorders
Chairs: Walton T. Roth & Steven H. Woodward
Participants: Bruce N. Cuthbert, Alex L. Gerlach, Frank H. Wilhelm, Steven
H. Woodward
Discussant: Walton T. Roth
Because of the prominence of both episodic and chronic hyperarousal symptoms
in the anxiety disorders, these disorders have been a focus of clinically
oriented psychophysiology for many years. Since all animals have fear systems,
infrahuman neurophysiological research into fear has pushed deep into the
brain; probably more is known about fear's biological substrates than about
the substrates of any other domain of psychopathology. One would expect,
therefore, that clinical application of psychophysiology to the anxiety
disorders would be particularly highly developed and could be considered
an important test case for the clinical utility of psychophysiology in general.Participants
will present research findings demonstrating the contributions as well as
limitations of psychophysiological methods as applied to the diagnosis,
management, and theoretical understanding of anxiety disorders such as simple
phobia, social phobia, agoraphobia, panic, and PTSD. Methods described will
include fear-potentiated startle, ambulatory cardiac, respiratory, and electrodermal
recording, facial blood volumetry for measuring blushing, and trauma-cue
induced responsivity assessed via multiple indices.We intend to address
these questions: What does psychophysiology have to offer the anxiety disorders
clinic, and what can it not yet provide? Where are clear avenues for progress,
and where are questions still ill-formed? What will be the role of ambulatory
psychophysiology? What are the limits of animal models of human fear in
explicating clinical anxiety?
Specific phobia and other anxiety disorders
Bruce Cuthbert, Evelyn Sullivan, Cyd Strauss, Margaret Bradley, & Peter
Lang
University of Florida
The clinical studies to be described are theoretically driven by basic research
on emotion, reflecting animal experiments involving states of appetitive
and defensive motivation. This work has defined a subcortical fear circuit,
including the sensory thalamus, amygdala, and periaqueductal central gray.
Research has shown that the startle reflex and autonomic reactivity are
augmented in fearful animals, and that these responses depend on an intact
fear circuit. Work with non-clinical human samples in our laboratory has
demonstrated a corresponding startle potentiation, in fear conditioning,
fear imagery, and while subjects attend to unpleasant pictures or sounds.
In exploring anxiety disorders psychopathology, these results are paralleled
most closely by patients with specific phobias. Clinically, these patients
most closely resemble non-anxiety populations, e.g., they show lower scores
on measures of general psychopathology and depression. Specific phobic patients
respond with a pattern that suggests a tightly organized, and reliably elicited,
fear memory structure. They show larger autonomic responses to imagery of
their own phobic material than other anxiety disorders groups. Further,
their reactions are reliably predicted by such factors as trait and state
imagery vividness, and general distress; such relationships are much reduced
for patients with social phobia and panic disorder. Specific phobics also
show marked fear-potentiated startle to phobia-relevant stimuli (whether
imagery or pictures), again exceeding comparable responses of other patients.
Thus, specific phobia best exemplifies activity in the fear circuits elaborated
by animal models, and represents a standard against which to evaluate fear-related
responding in other patient groups.
Subtypes of social phobia
Alexander L. Gerlach1, Frank H. Wilhelm2, & Walton T. Roth3
1Stanford University, 2Philipps-University Marburg, 3PAVAHCS
In recent years social phobia has attracted a substantial amount of research.
Considerable effort has been made to discriminate various subtypes of this
disorder. The hope is to isolate subcategories and ultimately to develop
specifically tailored treatments. Unfortunately either the discrimination
of these subtypes has turned out to be unreliable or it has been unclear
whether differences are qualitative or simply quantitative. The most common
distinction, that between generalized vs. non-generalized social phobics,
has usually been shown to be quantitative. Generalized social phobics score
higher on standard social phobia questionnaires, report more anxiety during
behavioral tests, and are rated as more severely impaired by clinicians.
However, a few studies have found that compared to generalized social phobics,
discrete social phobics have higher heart rate reactivity during performance
situations but lower reported anxiety, a dissociation that suggests a qualitative
difference. We will present data that failed to confirm this finding: In
our sample of 30 social phobics heart rate was higher in generalized than
specific social phobics in two of three social tasks. A neglected physiological
measure of social phobia is superficial facial blood volume. Blushing is
a common complaint of social phobics, distinguishing them from agoraphobics,
for example. We will document a larger increase in facial blood volume in
an embarrassing situation for social phobics than controls. However, we
will also show that social phobics in our sample who were especially concerned
about blushing did not exhibit greater changes in facial blood volume than
social phobics not concerned about blushing.
Multi-channel physiological assessment of a simple phobia in the natural
environment
Frank H. Wilhelm1 & Walton T. Roth2
1Harvard University, 2Stanford University
We evaluated the feasibility of recording multiple physiological anxiety
indices outside the laboratory, testing specific anxiety theories in 14
flight phobics and 15 controls. Benefits of baseline adjustment and transformation
for all variables, and adjustment of heart rate (HR) by ventilation to give
"additional HR" were calculated using individual relations between
ventilation and HR during an exercise test. Additional HR was not different
between phobics and controls when leaving the hospital, but it was significantly
greater for phobics when entering the airplane, accurately reflecting their
increased subjective anxiety while physically active. In contrast, HR did
not differ between groups on either occasion. A comparison of effect sizes
of 13 physiological parameters during flight showed that additional HR was
the best anxiety index even when subjects were seated. Although respiratory
rate and minute volume, indicators of hyperventilation, did not distinguish
groups, phobics paused more during inspiration than controls. Phobics also
showed more skin conductance fluctuations and less respiratory sinus arrhythmia,
pointing to enhanced sympathetic and reduced cardiac parasympathetic activation
in phobics during flight. Skin conductance level was related to reported
sweating. Accuracy of diagnostic classification was higher for the single
best self-report measure, SUDS anxiety, than for additional HR (90% vs.
79%). We conclude that ambulatory monitoring of multiple physiological systems
in phobic situations can yield objective information about certain symptom
categories, but does not necessarily sharpen diagnostic categorization.
Psychophysiological assessment of PTSD
Steven H. Woodward
PAVAHCS
The induction of sympathetic arousal by reminders of trauma has been proposed
to be a key diagnostic criterion for PTSD, perhaps even the "gold standard"
against which other diagnostic methods should be validated. Published reports
have relied principally upon heart rate acceleration to index sympathetic
arousal, and to a lesser extent upon skin conductance. "False negatives",
i.e., patients who meet DSM-III/IV criteria for PTSD but who fail to exhibit
substantial heart rate accelerations when exposed to trauma reminders are
frequently observed. False negatives reached a level of approximately 30%
in a recently completed DVA Cooperative Study of over 1000 Vietnam combat
veterans. While some consumers of this research may conclude that most false
negatives over-reported their PTSD symptoms on interview, such conclusions
are premature in light of our continuing ignorance about the nature of long-term
adaptation to trauma andthe methodological limitations of current studies.
Indices of arousal and arousal regulation which have yet to be measured
in studies of trauma-cue reactivity include heart rate variability, T-wave
amplitude, blink rate, pulse transit time, and startle probe amplitude.
Only one study has examined EEG. We will present data comparing Vietnam
combat veterans who were trauma cue "responders" versus "non-responders"
as defined by heart rate accelerations. These groups exhibited significant
differences in their levels of high frequency heart rate variability for
baseline periods during which they were exposed to positively-valenced stimuli.
This finding in itself suggests that the powerful downregulatory influences
on the heart must be considered when assessing trauma cue reactivity by
heart rate. Further group comparisons and individual case reports will be
presented to illustrate the complex interpretive challenge facing those
who try to diagnose PTSD psychophysiologically.
Ten years after: Paul Obrist's legacy to cardiovascular psychophysiology
Chair: Paul Grossman
Participants: Jasper Brener, James E. Lawler, Alan W. Langer, Kathleen C.
Light, James A. McCubbin
Paul Obrist died ten years ago, yet his language is still heard in the words
and voices of cardiovascular psychophysiologists the world over. His concepts
remain beacons and warnings to psychophysiology and behavioral medicine
as these disciplines enter the twenty-first century. In the wild and woolly
1960's and 70's, Paul Obrist, almost single-handedly, transformed cardiovascular
psychophysiology from a side-show of psychology where simple cardiac measures
were primarily employed as putative markers of covert psychological processes.
Obrist's vision of psychophysiology was to expand the reaches of cardiovascular
physiology to behavioral and psychological domains that had been previously
neglected by academic physiologists and physicians. This meant not only
describing the unique hemodynamic responses evoked by discrete mental tasks
and emotional states, but also addressing the mechanisms and biological
function of varying circulatory adjustments to behavior. To achieve these
ends, Obrist embarked upon an ambitious, systematic series of human and
animal investigations, pioneered the use of pharmacological agents in psychophysiological
investigation, and developed an intriguing set of hypotheses linking the
genesis of essential hypertension to stress-related cardiovascular reactivity
and systemic overperfusion. These are merely a few of his accomplishments.
Among his many personal attributes, Paul was keen, down to earth, straight
talking, and unafraid of a good argument. He displayed an uncommon sense
of passion and was a loyal friend and mentor. In this symposium, five of
Obrist's closest colleagues and former students will present their views
of the legacy he left us.
Obrist's view of the behavioral-cardiovascular interaction
Jasper Brener
State University of New York at Stony Brook
This presentation will describe experiments by me and my students that were
inspired by Obrist's ideas regarding behaviorally related variations in
cardiovascular performance. The concept of "overperfusion" is
central to Obrist's theory about the genesis of essential hypertension.
Normally cardiac output is precisely scaled to the perfusion requirements
of the tissues. However, sympathetically mediated increases in myocardial
performance augment blood flow to levels that exceed metabolic demand. This
overperfusion is counteracted by homeostatic mechanisms including vasoconstriction
that restore normal rates of blood flow. However, both the elevated cardiac
output and the subsequent increase in peripheral resistance cause blood
pressure to rise and are potentially pathogenic. These ideas of Obrist stimulated
us to investigate sources of variation in heart rate (as an index of cardiac
output) observed in behavioral situations. Experiments on the relationship
between heart rate, striate muscle activity and oxygen consumption were
conducted with Wendell Goesling, Ethel Eissenberg, Susan Middaugh, Keith
Phillips, Sam Connally Andy Sherwood and Jennifer Moses. These studies indicated
that most behaviorally related variations in heart rate could be accounted
for by metabolic demand that, in turn, could be attributed to striate muscular
activity. However, under certain conditions, heart rate was elevated above
metabolic demand. According to Obrist, these suprametabolic elevations in
cardiac activity were also triggered by incipient motor processes - preparation
for action. This notion has been examined experimentally by Sherwood, Moncur
and me, and theoretically by Requin, Ring and myself. More recently, Obrist's
ideas have stimulated experiments by Roy and Sequeira-Martino that identify
brain sites of striate muscular-visceral interaction in cats.
The borderline hypertensive rat (BHR) as an animal model for studying
the mechanisms of environmentally induced hypertension
James E. Lawler & Ingrid P. Edgemon
University of Tennessee
While psychophysiologists are familiar with the contributions of Paul Obrist's
research with human subjects, they are less familiar with his contributions
utilizing animal subjects. His early work with dogs will be briefly reviewed.
It is this work that led to his interest in assessments of aspects of cardiovascular
function in humans other than heart rate. The dog work was thus an early
tool for his emphasis on becoming more biological in all psychophysiological
research. That orientation has had a profound impact on my own career, and
that impact will be briefly reviewed. The development of the borderline
hypertensive rat (BHR) will be discussed. The BHR is the first generation
cross of inbred rats with genetic hypertension and those with no positive
family history of hypertension. The BHR, but not the normotensive rat, shows
chronic elevations in blood pressure when subjected to shock-shock conflict
stress. These elevations persist even in the absence of continued stress.
Second, a high salt diet also produces a chronic elevation of blood pressure.
Third, an exercise program can prevent stress-induced hypertension in this
model. Once the model was developed, research was then directed at elucidating
the mechanisms involved. The basic orientation is that there are central/renal
interactions involved in producing permanent hypertension in the BHR. Recent
studies addressing the role of these mechanisms will be reviewed. These
data show a critical role for the renal nerves, especially during the developmental
phase of hypertension. CNS changes also show patterned alterations in the
BHR which are absent in the WKY, when using either compound stressors (salt
and stress) or salt intake alone. Finally, the BHR appears to be an ideal
model for studying the role of family history in altering cardiovascular
reactivity to stress following chronic interventions such as exercise or
high salt intake.
The legacy of the cardiac-somatic hypothesis: This one's for you, Paul
Alan W. Langer & Mark R. Larson
Syracuse University
Introduced by Paul Obrist in 1968, the cardiac-somatic hypothesis continues
to exert a remarkable influence on cardiovascular psychophysiological research.
Initially advanced to elucidate biological processes when seeking to establish
what Paul called "biological yardsticks of psychological states,"
this simple but elegant formulation went on to frame just about every significant
problem addressed by cardiovascular psychophysiology. The most significant
application of the cardiac-somatic hypothesis continues to be the problem
of cardiovascular reactivity. The role of the sympathetic nervous system
in mediating cardiac reactivity to behavioral stress has been repeatedly
shown to occasion a dissociation of cardiac and somatic processes, the best
examples of which are additional HR and overperfusion (elevation in cardiac
output in excess of metabolic demand). Perhaps the most vexing problem for
Paul was the relationship between overperfusion and blood pressure control
and its potential role in the development of essential hypertension. Paul
speculated that overperfusion might be relevant to conditions such as borderline
hypertension, but this was refuted by several earlier studies suggesting
that the elevation in cardiac output, at rest in borderline hypertensives,
was linked to an appropriate increase in metabolic demand (oxygen consumption).
Unfortunately, Paul never got around to systematically examining this himself.
Nevertheless, we would now like to present new data which shows that persons
with mildly elevated blood pressures not only exhibit overperfusion during
a baseline preceding exercise but that this effect is potentiated during
the baseline prior to an active coping stressor. The legacy of the cardiac-somatic
hypothesis continues. This one's for you, Paul.
The reactivity hypothesis: Building on an uneven foundation
Kathleen C. Light
University of North Carolina at Chapel Hill
Paul Obrist was a spokesman for the potential importance of high cardiovascular
(CV) reactivity to stress in the pathogenesis of CV disease. Some have suggested
that little additional predictive information is obtained beyond that provided
by traditional risk factors. Also, there is a need for refinement of the
reactivity hypothesis itself. It has been shown that increased exposure
to life stress, operationalized as job strain, lack of social support, etc.,
increases the risk of adverse CV outcomes. Thus, a high reactor should not
demonstrate the same subsequent BP if s/he spends the next 20 years in a
low stress vs. high stress context. The tests for the reactivity hypothesis
should examine outcomes as a function of both the independent (additive)
and interactive effects of high CV reactivity with measures of exposure
to life stress. In a recent 10-year follow-up of 103 male college students
initially tested by Obrist et al, current clinic SBP was best predicted
by a model in which SBP during a reaction time task plus current score on
the Daily Stress Inventory plus their interactions added to a base model
of traditional risk factors (model R2 = 28% vs. base model R2 = 15%). Among
58 of these men who also completed 24-h ambulatory monitoring at follow-up,
DBP at work alone was best predicted by DBP during the cold pressor, plus
current job supervisor support, plus their interaction, added to body mass
index (BMI) (model R2 = 31%), while DBP at home with others was best predicted
by DBP during the reaction time task plus total social support plus their
interaction, added to BMI (model R2 = 24%).
A neo-Obristian view of the contributions of Paul A. Obrist
James A. McCubbin
University of Kentucky
The late Paul A. Obrist was a man who infused into his work both a romantic
passion and a lunchbucket-style work ethic. Obristian psychophysiology,
as the forerunner of mechanistic behavioral medicine, relied heavily on
methodological development as well as conceptual clarity. Paul spent years
developing his measures, from indices of left ventricular function to his
automated blood pressure contraptions that anticipated much of today's remarkable
tools. He advocated multiple converging methodologies. For example, to verify
the accuracy of his blood pressure devices he used direct radial artery
cannulae. He assessed the sensitivity of these devices using amyl nitrite
and other heroic manipulations to produce reliable blood pressure changes.
Therefore, his research was grounded in an empirical foundation that maximized
the impact of his findings. Paul also pioneered practical approaches to
semi-quantitative assessment of autonomic tone in the intact behaving human.
The creative use of pharmacological receptor blocking drugs during behavioral
manipulations allowed Paul to implement a degree of biological specificity
heretofore unknown in our field. Finally, his appreciation of the predictive
utility of individual reactivity differences in the disease process has
probably stimulated more research than any other single contribution. While
Paul may not have been the first to embrace these varied methodological
and conceptual issues, he parlayed them into a research juggernaut that
pushed the field into a new era. Paul bore heavily the birth trauma of these
times. However, the Neo-Obristian conceptus flourished through the next
decade and continues to stimulate exciting new research.
ERPs to subliminal stimuli: Markers for unconscious processes
Chair: Howard Shevrin
Participants: Howard Shevrin, Scott Bunce, Edward Bernat, Philip Wong
Discussant: Marcel Kinsbourne
As cognitive psychologists have begun to agree that unconscious processes
exist, it becomes of increasing interest to ask if ERPs can provide markers
for these unconscious processes. This symposium will present evidence that
ERPs can provide neurophysiological markers for unconscious processes elicited
by subliminal stimuli. It will be shown that subliminal ERPs have a component
structure (N1, P2, P3, a late negative peak) similar to supraliminal ERPs.
In the studies to be reported, stimuli have been presented at the objective
threshold (d'=0), the "gold standard" in subliminal research.
The first paper will introduce the topic, provide a brief review of pertinent
research, outline the methodological issues, and examine several theoretical
implications for our understanding of conscious and unconscious processes.
The second paper will present data on subliminal P3, showing its similarity
to supraliminal P3 in terms of wave form, scalp topography, and as an index
of psychologically relevant stimuli. The third paper will report on the
relationships between ERP components N1, P2, P3, a late negative peak, and
the connotative meaning of words as measured by the Osgood Semantic Differential
in both subliminal and supraliminal ERPs. The fourth paper will present
evidence that P3 can serve as a marker of learning in subliminal and supraliminal
classical conditioning paradigms.
Subliminal ERPs: History, theory, methods
Howard Shevrin
University of Michigan
The earliest research evidence of ERPs to subliminal stimuli was reported
by Shevrin and Rennick (1967, Psychophysiology, 3, 381),Shevrin and Fritzler
(1968, Science, 161, 295), and Libet, et al (1967, Science, 158, 1597).
The first two studies demonstrated that a P2 component could discriminate
between two visual stimuli presented at 1 ms. The third study, employing
direct cortical activation, demonstrated that early somatosensory ERP components
(approximately 100 ms) were correlated with subthreshold stimuli. Subsequently,
Shevrin and co-workers published a series of studies replicating and extending
the original work summarized in Shevrin (1973, Psychological Issues, 30,
56). Other laboratories have reported additional findings (Kostandov and
Arzumanov, 1977, Acta Neurobiologiae Experimentalis, 37, 311; Barkoczi et
al., 1983, Psychologia, 26, 1); Brandeis and Lehmann, 1986, Neuropsychologia
24, 151). Libet et al (1967) argued that later endogenous ERP components
like P2 and P3 were associated with consciousness, a position also taken
by Posner and Boies (1971, Psychological Review, 78, 391). Results from
other laboratories (Shevrin, 1973; Kostandov and Arzumanov, 1977) presented
evidence that both P2 and P3 were found in subliminal ERPs. Subsequent research
to be reported in this symposium by Bunce, Bernat, and Wong have provided
further evidence for subliminal P2 and P3 components. These findings suggest
that unconscious processing can continue well below 300 milliseconds. ERPs
to subliminal stimuli allow investigators to study concomitant unconscious
processes rather than relying, as in most purely cognitive studies, on inferring
unconscious processes from indirect effects such as priming.
P300 to subliminal stimuli: Looks like a duck, acts like a duck...
Scott C. Bunce1, Edward Bernat1, Howard Shevrin1, & Stephen Hibbard2
1University of Michigan, 2Pacific Graduate School of Psychology
Findings will be presented demonstrating that one endogenous ERP component,
P300, can be used for the study of complex unconscious psychological processes.
Three lines of evidence will be presented that argue for the utility of
P300 in subliminal paradigms. First, there is a characteristic waveform
that occurs approximately 300 ms post-stimulus, which will be illustrated
through three-dimensional plots and grand averages. Second, when task and
stimuli were identical across subliminal and supraliminal conditions, the
P300 amplitude produced in a subliminal condition had the same scalp topography
across electrodes (F3, F4, CzPz, P3, P4, Oz) as the P300 produced in the
supraliminal condition (rank order correlation = 1.00). Third, P3 amplitude
and latency can be used as an index of psychologically relevant stimuli.
Research will be presented demonstrating that P300 can index individual
differences in response to identical words, both as a function of traumatic
life experience (early parental loss) and of subliminal versus supraliminal
stimulus presentation. On the basis of previous research (Bunce et al.,
1996, Psychophysiology, 33, S26), it was predicted that Loss participants
would experience positive and intimate words as more negative than Controls,
but only when presented outside awareness. P300 amplitude differentiated
between positive and negative mood adjectives at supraliminal durations,
but did not distinguish Loss from Control groups. At subliminal durations,
however, a significant interaction indicated that whereas Control participants
showed a greater P300 response to the negative words, P300 failed to discriminate
between positive and negative words for the Loss group.
Subliminal and supraliminal ERPs as markers of connotative meaning
Edward Bernat1, Scott Bunce1, Howard Shevrin1, Michael Snodgrass1, &
Steve Hibbard2
1University of Michigan, 2Pacific Graduate School of Psychology
Findings are presented demonstrating that visual ERPs elicited by subliminal
and supraliminal stimuli differentiate pleasant and unpleasant connotative
meaning as measured by the Osgood Semantic Differential Scale. Correspondence
in structure between ERPs elicited at subliminal and supraliminal stimulus
durations are demonstrated by comparing grand average plots. Two experiments
are presented which demonstrate that ERP components (N1, P2, P3, a late
negative peak) can discriminate between individually rated pleasant and
unpleasant stimuli at both subliminal and supraliminal durations. In the
first experiment with normal participants, a positive amplitude shift occurred
across all components for unpleasant stimuli, at both subliminal (F[4,64]=2.50,
p<.05) and supraliminal (F[4,64]=3.40, p<.02) durations. In the second
experiment, involving social phobics, an interaction between stimulus duration
and connotative meaning was found: supraliminally, a positive amplitude
shift occurred for unpleasant words, while subliminally a positive amplitude
shift occurred for pleasant words (F[4,64]=14.18, p<.001). Differences
in subliminal response for the two samples are discussed in the light of
the fact that participants in the second experiment were social phobics.
These data demonstrate the similarity in structure of ERPs to subliminal
and supraliminal stimuli, and show that standard ERP component measures
can be an effective approach for investigating connotative meaning using
supraliminally and subliminally presented stimuli.
ERP indices of unconscious processes in classical conditioning
Philip S. Wong1, Scott C. Bunce2, Edward Bernat2, Howard Shevrin2
1New School for Social Research; 2University of Michigan
We report new findings from two previous studies in which we found evidence
for ERP correlates of unconscious processes involved in classical conditioning.
Specifically, we report new P3 findings consistent with results reported
by Bunce and Bernat (this symposium). In the first study (Wong et al., 1994,
Psychophysiology, 31, 87), conditioning was established supraliminally and
the effects measured subliminally. In the second study (Wong et al., submitted,
1997), conditioning was established subliminally and the effects measured
supraliminally. In both studies we used a paired-stimulus paradigm with
facial schematics, one unpleasant and one pleasant; the unpleasant face
was linked to an aversive shock. We found that in the first study P3 varied
as a function of face and conditioning. A significant interaction was present
between pre-conditioning and post-conditioning subliminal durations across
Cz, P3, P4, and Oz for P3 amplitude, F(1, 14) = 5.52, p<.03. A trend
was also found for P3 latency for the same interaction, F(1, 15) = 3.44,
p = .08. The direction for the amplitude interaction indicated that for
the CS- habituation had occurred (F[1, 14] = 3.50, p<.08), while no habituation
had occurred for the CS+. Latency increased for the CS+ from pre- to post,
and decreased for the CS-, but neither effect was significant. These results
are consistent with P3 results from the second study in which habituation
occurred for the CS+ from pre-to post supraliminal durations. Thus, P3 demonstrated
the effects of conditioning both sub- and supraliminally.
Review and funding of psychophysiological research: Perspectives,
opportunities, and strategies
Chair: Cindy M. Yee-Bradbury
Participants: Richard K. Nakamura, Niels Birbaumer, Margaret M. Bradley,
Gregory A. Miller
Research funding can be critical to a program of psychophysiological research.
The goal of this panel discussion is to familiarize researchers with current
perspectives and judgments concerning the scientific review and support
of psychophysiological research. Members of the panel will present prevailing
views on grant applications, the review process, funding opportunities,
and the prospects of receiving funds to support psychophysiological research.
Perceptions as well as common misperceptions about psychophysiological research
also will be discussed. Various perspectives are represented on the panel,
ranging from that of funding agencies to those of experienced grant reviewers.
Following presentation of key issues by members of the panel, the discussion
will be opened up for questions and comments from the audience.
Perspectives on current restructuring of review at the National Institutes
of Health and funding priorities of the National Institute of Mental Health
Richard K. Nakamura
National Institute of Mental Health
Many aspects of the mechanisms for federal funding of research in the United
States are undergoing review and reorganization. Dr. Richard Nakamura, Director
of the Office of Science Policy and Program Planning at the National Institute
of Mental Health (NIMH), will discuss the merger of NIMH study sections
into the National Institutes of Health (NIH). An overview of the newly reconfigured
study sections, and their relevance to psychophysiological research will
be provided. Dr. Nakamura also will discuss the implications of this merger
for NIH grant applications and review. He will address the development of
NIMH funding priorities, and the criteria by which applications are judged.
In addition, the perceived value and limitations of psychophysiological
research, from the perspective of NIMH, will be considered.
Recent trends in psychophysiological research funding in Europe
Niels Birbaumer
University of Tuebingen
Multiple mechanisms exist for obtaining funds to support psychophysiological
research in Europe. Drawing from experience with the European Community,
Human Frontiers, German Research Society, and National Ministries in Europe,
Prof. Dr. Birbaumer will discuss the review process and provide practical
information that will maximize chances for obtaining funding. Comparisons
will be made between various funding agencies. The likelihood of obtaining
funds for psychophysiological research also will be considered.
Show me the $$$! Tips for preparing grant applications
Margaret M. Bradley
University of Florida
Whereas granting agencies typically provide specific guidelines for the
preparation of a research application, certain additional features that
can be important in the peer review process are not always explicitly mentioned
and in fact can sometimes seem contradictory (e.g., achieving clarity and
conciseness within a comprehensive and detailed discussion). Serving on
the review side of the granting process allows one to become more familiar
with different methods for successfully balancing these apparently conflicting
objectives and for preparing a successful application. In addition, whereas
the relationship between reviewer and investigator is sometimes viewed as
adversarial, serving on a National Institute of Mental Health initial review
group (IRG) committee also makes it apparent that the reviewer often acts
as the investigator's advocate in the review process. In this case, the
success of presenting and defending a research proposal to the larger committee
is dependent to a large extent on the information available in the grant
application. Thus, the investigator's task is not only to clearly describe
and justify the proposed research but also to provide the reviewer with
as much information as possible that will assist in supporting the application
in the larger committee meeting. These and other suggestions gleaned from
IRG service will be discussed in this session.
Outsiders inside: Some notes on Initial Review Group culture
Gregory A. Miller
University of Illinois at Urbana-Champaign
The experience of serving on a National Institute of Mental Health Initial
Review Group and reviewing federal grant applications is valuable in many
ways. One reads state-of-the-art literature reviews that are more or less
in one's area, one watches paradigms in the making, one learns how to read
and thus how to write a grant application, and one is paid (precious little!)
to argue constructively with smart colleagues about the merits of proposed
protocols and even research areas. Accompanying this intensive, advanced
seminar experience are some windows into how federal funding agencies work,
how they see themselves, and how they see their various constituencies.
Experiences and perspectives from serving on several review groups will
be shared.
Psychophysiological studies of biobehavioral mediators of risk
for alcohol abuse
Chair: Peter R. Finn
Participants: Alan R. Lang, Jeanette Taylor, Howard L. Cohen, Peter R. Finn
Discussant: Kenneth J. Sher
Alcoholism is a heterogeneous disorder with a number of associated risk
and vulnerability factors such as a positive family history of alcoholism,
antisocial/disinhibited behavior, and differential responsivity to alcohol.
Although it is clear that each of these factors is associated with increased
risk for the disorder, the precise mechanisms by which they mediate risk
are unclear. Current etiological theories are typically biobehavioral in
nature and focus on the role of individual differences in vulnerability
(e.g., a genetic diathesis) in the development of substance abuse. Psychophysiology
provides a set of methods and measures that can open a unique window for
identifying biobehavioral mediators of risk for alcohol abuse. The purpose
of this symposium is to present data from range of psychophysiological studies
of biobehavioral mediators of risk with an emphasis on the interface between
substance abuse and antisocial/disinhibited behavior. The symposium includes
reports of studies that include autonomic, somatic, and electrophysiological
measures and paradigms representing a range of approaches to investigating
risk factors in substance abuse.
Drinking, emotional response, and disinhibition: Startle-probe studies
implicating complex cognitive-affective processes in the connections.
Alan R. Lang & Christopher J. Patrick
Florida State University
A multidimensional conceptualization of emotion incorporating multilevel
brain processes provides enhanced opportunities for understanding connections
between alcohol intoxication and emotional response, and perhaps behavioral
disinhibition that sometimes accompanies them. Three investigations of drugs
and affect, all incorporating startle-probe methodology to index emotional
valence, highlight the importance of multidimensional assessment of emotion
by yielding findings at variance with those emanating from earlier, arousal-focused
research apparently supporting the thesis that alcohol intoxication is routinely
accompanied by tension reduction or a dampening of response to stress. Key
results indicate that, whereas diazepam selectively reduces startle potentiation
associated with aversive stimuli, alcohol merely reduces overall reactivity
without having a significant impact on affectively modulated startle. Furthermore,
in a study of how alcohol and pleasant distracters affect startle reactions
under threat conditions, diminished startle potentiation was observed only
in the presence of both alcohol and distraction. This suggests that the
compromising of higher-order cognitive processes (e.g., those involved in
simultaneous attention to competing stimuli of differing salience) by alcohol
may be critical to any stress reduction associated with drinking. Speculative
extrapolation from these findings to the conditions that may mediate alcohol-related
behavioral disinhibition (e.g., aggression) are intriguing and may have
some parallels in distinctions between classic psychopathy and simple antisociality.
Electrodermal response modulation: Could it index a protective factor against
substance dependence in adolescent males?
Jeanette Taylor, Scott R. Carlson, William G. Iacono, & David T. Lykken
University of Minnesota
Adolescent males may exhibit a number of disinhibited behaviors including
antisocial behavior and substance use. The development of substance dependence
among male adolescents may result from a combination of factors including
persistent substance use despite predictable aversive consequences. Poor
electrodermal response modulation to a predictable aversive stimulus may
index a constitutional factor related to the development of substance dependence.
As such, we expected male adolescents with poor electrodermal response modulation
to a predictable aversive stimulus to have more symptoms of DSM-III-R substance
dependence than those with good modulation. As part of the ongoing Minnesota
Twin Family Study, we recorded skin conductance responses (SCRs), heart
rate (HR), and anticipatory skin conductance level (SCL) and non-specific
electrodermal fluctuations (NSFs) from 75 16- to 18-year-old males during
five 100-s trials in which a 2-s 90 dB white noise blast was either unpredictable
or temporally predictable. The modulation index reflected the percent increase/decrease
in SCR when the stimulus was made predictable. The 75 boys fell into three
equal-sized modulation groups. Good modulators had significantly fewer symptoms
of Alcohol and Nicotine Dependence than Moderate and Poor modulators. Mean
anticipatory SCL did not differentiate the modulation groups, but Good modulators
had significantly more anticipatory NSFs than Poor modulators. Finally,
mean HR did not differentiate the modulation groups. Given the similarity
in symptom counts between the Moderate and Poor modulators, it could be
that good SCR modulation indexes a protective factor against substance dependence
in adolescent males.
Electrophysiological correlates of disinhibition in individuals at high
risk to develop alcoholism
Howard L. Cohen, Bernice Porjesz, & Henri Begleiter
SUNY Health Science Center-Brooklyn
For the past twenty years our laboratory as well as several others has repeatedly
observed a significantly lower amplitude of the P3 component of the Event-Related
Potential (ERP) in both abstinent alcoholics and the offspring of alcoholics.
It has been demonstrated that this important finding is not the result of
excessive alcohol abuse but is highly related to family density of alcoholism.
We have recently postulated that this significantly reduced P3 component
may indicate increased central nervous system (CNS) disinhibition. Loss
of inhibitory control is not only prevalent in abstinent alcoholics, but
also is present in children of alcoholics. It is well-established that male
offspring of alcoholics are more likely to display undercontrolled behavior,
and children with conduct disorders are more likely to have a positive family
history of alcoholism. Thus, there is evidence that disinhibitory processes
may underlie both electrophysiological deficits and behavioral dysfunction
in alcoholics and their offspring. In our laboratory we have been assessing
ERP correlates of disinhibition with novel neurophysiological paradigms.
The results of these electrophysiological studies suggest that aspects of
brain dysfunction (i.e. lack of differential inhibition) may be involved
in a predisposition for alcoholism.
Biobehavioral mediators of early-onset, antisocial, alcohol abuse in
men and women: Response to threat and passive avoidance learning
Peter R. Finn, Linda Rorick, & Joseph Steinmetz
Indiana University
Substantial evidence exists for the validity of an early-onset, antisocial,
subtype of alcoholism. Early-onset alcoholism is associated with severe
symptoms, poor treatment response, more alcoholic relatives, impulsivity,
novelty seeking, and low harm avoidance (Cloninger, 1987). Studies indicate
that low harm avoidance and high novelty seeking in young boys are predictive
of adolescent-onset alcohol abuse. This literature presents 2 major issues
and questions: (i) what is the role of gender, and (ii) what biobehavioral
mechanisms mediate risk for this type of alcohol abuse. Much of the research
is focused on males, and relies on self/behavioral report of symptom phenotypes.
This study assessed the role of disinhibitory processes, as evidenced by
low electrodermal (EDA) and fear-potentiated startle response to threat
and deficits in passive avoidance learning, in early-onset (less than 20
years), antisocial alcoholism. The data revealed that early-onset alcoholism
was associated with high levels of antisocial traits, low harm avoidance
(in men more than women), increased venturesomeness and impulsivity, and
a family history of alcoholism. EDA hyporeactivity was associated with low
harm avoidance, a family history of alcoholism, and early-onset alcoholism.
Decreased fear-potentiated startle was related to low harm avoidance and
increased venturesomeness but only in men. In men, but not women, deficits
in passive avoidance learning were also associated with antisocial traits,
a family history of alcoholism, early-onset alcoholism, EDA hyporeactivity,
and low harm avoidance. In women, early-onset alcoholism and antisocial
traits were associated with good passive avoidance learning. Results are
discussed in terms of a disinhibitory predisposition to alcoholism and potential
gender differences in phenotype manifestation.
Synthetic Emotion
Chair: Louis G. Tassinary
Participants: Rosalind W. Picard, Frederic I. Parke, Stephen Grossberg
Discussant: Louis G. Tassinary
Synthetic emotions are a perennial topic of conversation in philosophy and
science fiction. The fields of affective computing, character animation
and the mathematical modeling of cognitive and neural systems have recently
become sufficiently advanced to where it is possible to broach the synthesis
of emotion in a scientifically credible manner. Not unlike how the use of
model systems and computational neuroscience have fundamentally changed
cognitive science, advances in these fields are poised to change the nature
of affective science. Our three speakers will provide reports on their extensive
forays into each of these areas. First, Picard will describe recent work
on attempts to make computer-laden environments more "human-like"
though the addition of novel sensors and effectors, as well as through acquiring
"emotion-like" states. Second, Parke will review the state-of-the-art
in computerized facial animation, an area at the forefront of computerized
character animation. Finally, Grossberg will provide a tutorial on a type
of model for adaptive systems generally that may provide the foundation
for understanding and ultimately constructing autonomous synthetic agents
that react "emotionally" and learn from experience. Following
discussion by the chair, questions and comments from the audience will be
invited.
Affective computing
Rosalind W. Picard
Massachusetts Institute of Technology
Affective computing is computing that relates to, arises from, or deliberately
influences emotions. There is evidence that computers with emotion-like
mechanisms could be more flexible and rational in tasks such as planning
and decision making than could non-affective machines, so that soon computers
may also be given the ability to "have emotions", or at least
emotion-like mechanisms. Computers are also beginning to acquire the ability
to express emotions and to recognize emotional expressions. It is known
that emotions are a natural part of human-computer communication; consequently,
it is increasingly important for computers to learn how to perceive the
emotions a user is expressing. For example, software agents that try to
learn a user's preferences could be less of a burden to their users if they
could recognize whether a user liked or disliked their most recent action.
We are developing wearable computers, equipped with physiological sensors,
which can be in long-term personal contact with their wearer, learning how
he or she responds in a variety of natural situations. We are also developing
pattern recognition algorithms that can be applied to combinations of physiological
signals in an effort to find patterns consistent with various emotional
states. The plan is that given physiological patterns, together with high-level
knowledge about the user's situation, then a computer could learn to recognize
his or her probable emotions. I will present our research in this new area,
focusing on sensing, signal processing, and pattern recognition problems
as well as on some emerging applications.
Computer facial animation
Frederic I. Parke
Texas A&M;University
Facial animation is a major aspect of computer based character or "actor"
animation. Facial animation is also a part of recent user interfaces which
use anthropomorphic agents or avatars. In both uses, capable facial animation
must be able to express and convey "synthetic emotion." Development
of computer facial animation began in the early 1970's, with most development
occurring in the last decade. Facial animation is based on the development
of geometric models for the various facial features. These are based on
the anatomy of the face - facial skeleton, muscles, skin and interior structures
of the mouth such as the tongue. Facial models are often concerned with
capturing the likeness of specific real faces. This is done by digitizing
the shape of real faces using one of several three-dimensional surface measuring
techniques. Animation is accomplished by changing facial shape over time
using interpolation, muscle based deformation, or surface parameterization
techniques. Facial models usually incorporate the mechanics of facial tissue.
They can also include vascular skin color effects. Successful facial models
support a wide range of expressions including the universal expressions:
surprise, fear, disgust, anger, happiness, and sadness. Speech synchronized
animation involves matching speech postures to a spoken soundtrack. Complete
speech animation includes coarticulation, expression overlays and eye actions.
Current goals in facial animation research include the development of models
with expression abstraction layers, higher level controls, and ultimately
synthesized behaviors and personalities.
How do reward, emotion, and attention help to shape brain decisions?
Stephen Grossberg
Boston University
This talk will provide a self-contained introduction to neural models that
are being developed to explain how cognitive and emotional processes interact
in the brain. These models clarify how cognitive-emotional interactions
help to focus attention on important objects and events, and influence the
affective decisions that we make in our daily lives. The results clarify
a large body of psychophysiological data concerning classical and instrumental
conditioning, decision making under risk, and various mental disorders,
including examples of schizophrenia, juvenile hyperactivity, depression,
and Parkinson's disease. A key theme in the models is how opponent processes
help to control reinforcement learning in response to positive and negative
reinforcers. Affective antagonist rebounds are triggered in these opponent
circuits when reinforcers are suddenly removed, or when other novel events
occur. These rebounds can be traced to the action of chemical transmitters
that are slowly inactivated, or habituate, in an activity-dependent way.
Inverted U properties are also found that help to differentiate symptoms
of overaroused vs. underaroused depression. Paradoxical data concerning
decision-making under risk, including the gambler's paradox and preference
reversals naturally emerge from these circuits as emergent properties. Cognitive-emotional
interactions that include these opponent circuits clarify how learning can
distinguish between contingencies that are causal, vs. accidental correlations
in the environment. In particular, it is shown how unexpected consequences
can be used to actively extinguish otherwise long-lasting memories. The
model circuits are interpreted in terms of thalamo-cortico-amygdala-hippocampal
interactions in the brain.
Steady-state evoked potentials: New methodological approaches
and applications
Chair: Matthias M. Mueller
Participants: Terence W. Picton, Scott Makeig, Livio Narici, John W. Rohrbaugh,
Matthias M. Mueller
The steady-state response (SSR) is a continuous brain response that is elicited
by a repetitive stimulus presented at a fixed frequency. The SSR can be
recorded from the scalp as a nearly sinusoidal oscillatory waveform having
the same fundamental frequency as the driving stimulus and often including
higher harmonics. Steady-state responses have been reported in the acoustic,
visual and somatosensory modalities. Up to today the mechanism underlying
the generation of steady-state responses is still under debate. However,
steady-state responses serve as a important tool to investigate normal and
abnormal visual processing and cognitive processes including attention,
memory, and decision making. This symposium will discuss the latest methodological
approaches to analyzing and localizing theses responses, to gain more information
on the underlying mechanisms. Interactions between separate brain regions
will also be discussed by means of steady-state responses, and clinical
applications will be demonstrated. In addition, it will be shown that high
frequency visual evoked steady-state responses are modulated by spatial
attention, indicating that steady-state responses serve as an useful tool
to investigate cognitive processes. Given the increasing interest in research
on steady-state responses, and the growing evidence that SSRs provide important
information for understanding basic brain functioning, this symposium will
provide an opportunity to demonstrate and critically discuss (a) the progress
the field has gained so far and (b) avenues for research in the future.
Studying cerebral function with steady-state responses
Terence W. Picton
University of Toronto
Steady-state evoked potentials possess several features that make them useful
tools for evaluating cerebral function. First, unlike temporal waveforms,
steady-state responses do not require the experimenter to identify peaks
prior to measurement. The responses can be easily and objectively measured
in terms of the real and imaginary components of the response (or amplitude
and phase) at the frequency of stimulation or some harmonic thereof. These
two-dimensional measurements can be distinguished from noise using well
defined statistics and can thus be used as objective indices of sensory
activation. Since the real and imaginary components are independent, they
can be directly submitted to source analysis, which shows sources that are
mainly specific to the modality of the stimulus. Second, steady-state responses
can be continuously and consistently recorded over intervals lasting from
seconds to minutes. This allows easy comparison to measurements of cerebral
blood flow which have a similar time-course. These temporal characteristics
also allow steady-state responses to monitor fluctuations in cerebral activity
related to arousal or attention. Auditory 40 Hz responses are particularly
susceptible to anesthesia and can be used to monitor anesthetic unconsciousness.
Selective attention can be studied using two or more trains of stimuli,
each of which can be attended or ignored, and each of which can be analyzed
separately. Third, the steady state responses may be able to evaluate various
rhythmic interactions in the human brain by driving rhythms that are important
to cognition but too evanescent to study during normal activation.
Independent component analysis of steady-state responses
Scott Makeig1,2, Tzyy-Ping Jung1,3, Te-Won Lee3, & Terrence J. Sejnowski3
1Naval Health Research Center-San Diego, 2University of California-San Diego,
3The Salk Institute-La Jolla
Independent Component Analysis (ICA) is a new signal processing technique
for decomposing spontaneous or evoked EEG and MEG data into temporally independent
and spatially fixed components. The scalp distribution of the auditory steady-state
response near 40 Hz appears to sweep from the front to the back of the scalp
every cycle. ICA decomposes this apparent movement into the sum of at least
two bilateral components with different scalp distributions and phase lags.
ICA accounts for the transient perturbations in SSRs produced by experimental
events using the same components producing the SSR, supporting the hypothesis
that these transient (CERP) perturbations represent modulation of the ongoing
response. Application of ICA algorithms capable of both sub-Gaussian and
super-Gaussian components will be presented and psychophysiological implications
of new blind decomposition techniques discussed.
Repetitive sensory stimulation and cortical rhythms: Neuromagnetic studies
Livio Narici
University of Rome
Cortical responses to repetitive sensory stimulation are studied by delivering
a short train of stimuli separated by pauses. They are analyzed both during
stimulus delivery and pauses, while scanning the stimulus rate within the
frequency band of interest. This "Frequency Responsiveness Procedure"
was applied to healthy subjects during two different investigations, in
Rome (CNR-IESS, 28-channel neuromagnetic sensor, visual and somatosensory
modalities) and in Helsinki (LTL-Brain Research Unit,122-channel whole scalp
neuromagnetometer, visual somatosensory and auditory modalities).Electric
0.3-ms pulses were delivered to the median nerve at an intensity above motor
threshold. 30-ms light pulses were presented with LEDs at 5-7 degrees eccentricity
in the left hemifield. 15-ms monaural noise bursts were used for the auditory
modality. Stimuli were delivered in separate sessions in trains of 14-15
stimuli, interleaved with 1.3-1.6 s pauses. Stimulus rates within the trains
varied from 6 to 14 Hz.Three effects were observed in relation to repetitive
stimulation: i) activity driving (modality specific); ii) rhythm suppression
(also in non-modality specific areas); iii) minor but consistent wide band,
activity suppression over the whole cortex. Different activities present
in the same frequency band were discriminated on the basis of their responsiveness
to repetitive stimulation. The genesis of the measured visually driven activity
(composition of evoked responses vs. driving of the intrinsic rhythms) appeared
to be related to the amount of posterior resting rhythm. Source localization
was performed whenever possible.
Event-related perturbations in steady-state rhythms
John W. Rohrbaugh, Erik Sirevaag, & Andrei Vedeniapin
Washington University School of Medicine
Event-related perturbations in steady-state auditory rhythms have been described
by a number of investigators as signs of transient cognitive and attentional
processes. The perturbation response is of particular interest because it
can be obtained nonobtrusively in response to probe stimuli under conditions
in which it does not interfere with performance of primary tasks. These
findings and interpretations will be reviewed. Data from a study of the
origins of the auditory steady state rhythm will presented, as will speculations
regarding the character of the event-related perturbations in rhythm phase
and amplitude. These findings are consistent with an interpretation of the
rhythm in terms of hybridized middle latency components, with lability of
the P1 response contributing principally to the perturbations of the rhythm.
Data from a study of chronic alcoholics will be presented to illustrate
possible clinical applications of the phenomena.
High frequency steady-state visual evoked potentials are modulated by
spatial selective attention
Matthias M. Mueller1, Terence W. Picton2, & Steven A. Hillyard3
1University of Konstanz, 2 Rotman Research Institute-North York (Canada),
3University of California-San Diego
To investigate the attentional modulation of high frequency steady-state
visual evoked potentials (SSVEPs), subjects attended to one of two "bars",
each of which consisted of a linear array of five bicolor (red, green) LEDs
flickering in the left (at 20.8 Hz) and right (at 27.8 Hz) visual hemi-field.
1296 ms after flickering onset, an attention-directing cue informed the
subject whether to attend to the left or the right bar. Subjects were instructed
to fixate on a central point and to detect targets within the bar that was
indicated by the cue for a period of 10 s per trial by pressing a button.
Targets were defined as a change in color of the most upper and most lower
LED. EEG was recorded from 30 scalp electrodes. SSVEPs to the concurrent
left and right field flickers were averaged separately in the time domain
by using a moving window technique. Amplitude was obtained by means of a
FFT at the appropriate frequency. Overall SSVEP amplitudes were larger for
the 20.8 Hz flicker than the 27.8 Hz flicker (p <0.001). With attention,
the SSVEP amplitude increased significantly for both the right and the left
bar (p <0.01). This increase was greater over the contralateral hemisphere
(p <0.05), and moreso for posterior than anterior electrode sites (p
<0.05). The present results suggest that SSVEPs might serve as an useful
tool to study visual attention.
The role of cortisol in memory, attention, and emotion
Chair: Louis A. Schmidt
Participants: Sonia Lupien, Melissa Goldberg, Louis A. Schmidt
Discussant: Jay Schulkin
The purpose of this symposium is to bring together researchers from disparate
literatures in order to better understand the multiple roles of cortisol
in cognitive and emotional processes in humans. We will address several
questions: 1) does elevated cortisol impair memory recall for specific classes
of stimuli (e.g., neutral objects) and enhance memory recall for other classes
of stimuli (e.g., fear-related objects)? 2) what are the functional aspects
of elevated cortisol in particular personality profiles (e.g., fearful and
shy individuals)? and 3) what are the developmental implications associated
with current work in this area in relation to the aging brain and cognitive
and emotion processes over the lifespan? Sonia Lupien will present data
on the relation between elevated cortisol and memory performance (implicit
and explicit) in young and elderly human adults. Melissa Goldberg and Nathan
Fox will present data in adults on the different effects of cortisol on
memory and attention. Louis Schmidt and Nathan Fox present data from studies
of children in which elevated endogenous cortisol levels are linked to fear-potentiated
startle responses and shyness during the preschool and early school age
years. They also present data on the relation between exogenous administration
of high doses of cortisol and the startle response, and they speculate on
the function of elevated cortisol in shy and fearful children and adults.
Jay Schulkin, known for his work on the effects of cortisol on cognitive
and affective processes, will integrate and facilitate discussion on the
three papers.
Cortisol and memory
Sonia Lupien
Douglas Hospital/McGill University
We previously reported that 38% of healthy elderly adults show a significant
year-to-year increase in cortisol levels, while 47% present a moderate year-to-year
increase and 16% present a year-to-year decrease of cortisol levels with
years. We reported a significant negative correlation between increase in
cortisol levels with years and decreased explicit memory performance. Implicit
memory was not correlated with increase in cortisol levels in this population.
Confirmation of the detrimental effects of increased levels of corticosteroids
on hippocampal function in this population was recently obtained in studies
showing that elderly subjects presenting a significant year-to-year increase
in cortisol levels with explicit memory impairments show a significant 17%
reduction of hippocampal volume as measured by magnetic resonance imaging
when compared to the other elderly subjects. We now report data on 8-year
cortisol secretion in 12 healthy elderly human subjects, and show that the
function relating cortisol secretion and time in these elderly subjects
is not linear as originally postulated, but rather fits a curvilinear function
with a three-year wave-cycle of cortisol secretion that slowly increases
with each cortisol/year peak. Analysis of 4-year cortisol secretion and
explicit/implicit memory function in other elderly subjects further showed
that the curvilinear increase observed in cortisol secretion in later life
is closely related to a curvilinear decrease in explicit memory function.
This suggests that there is some degree of reversibility in the cognitive
deficits generally reported in aged human populations.
Effects of cortisol on memory and attention
Melissa Goldberg & Nathan A. Fox
University of Maryland-College Park
We investigated in normal human adults the cognitive effects of prednisone
(160 mg daily; n=9) versus placebo (n=6) in a double blind procedure. Treatment
lasted for 4 days. On day 4, subjects performed two tasks: attention and
memory. (1) In a visual attention task, subjects indicated which of two
shapes appeared in the periphery 400 msec after a central cue, with those
shapes surrounded by compatible or incompatible distractors. (2) In a memory
task also on day 4, subjects studied a tray of 25 items for 1 min and then
had 2 min to recall the items; on day 9, subjects were asked to recall the
items presented on day 4. Results on the attention task showed the groups
did not differ with respect to the effects of the cue and/or the distractors
(ANOVA on RT, interactions with group, all p's >.10). This suggests that
prednisone treatment may not interfere with the abilities to shift attention
or maintain focused attention. On the memory task, although subjects in
the prednisone group were initially (day 4) able to remember as many items
as controls (12.8 versus 13.3, respectively, t-test, p > .10), 5 days
after the treatment ended (day 9), they were impaired in their ability to
remember as many items (7.9 versus 11.1, t-test, p <.001). The results
to date agree with reports in the literature that brief prednisone treatment
interferes with delayed recall information learned during treatment and
not visual attention.
The role of cortisol in fear-potentiated psychophysiological responses
in human adults
Louis A. Schmidt & Nathan A. Fox
University of Maryland-College Park
A number of studies of animals and humans suggest that cortisol may play
a role in the maintenance of fear responses. The purpose of this paper will
be to present data from a series of developmental studies in which the relation
between elevated endogenous and exogenous cortisol and fear responses have
been examined. In study 1, we found a significant relation between shyness
and elevated morning salivary cortisol. Shy preschoolers displayed high
morning salivary cortisol levels compared with their nonshy counterparts.
In study 2, we found modest relations between high salivary cortisol exhibited
in the laboratory and the fear-potentiated startle response in middle childhood.
Seven year-olds who displayed high laboratory salivary cortisol levels exhibited
a heightened startle amplitude in response to an affective social challenge
task. In study 3, we administered high doses of prednisone (160 mg) daily
for four consecutive days to 12 undergraduate males and a placebo to 12
undergraduate males in a double blind study. On day four, we presented the
Lang slide paradigm to the 24 subjects and examined their fear-potentiated
startle responses while viewing the slides. These data will be presented
and discussed in relation to previous findings from our laboratory on the
role of cortisol in the regulation of emotion.
Psychophysiology: The Shape of its Future
Chair: Emanuel Donchin
Participants: Richard J. Davidson, Emanuel Donchin, Edward S. Katkin, Peter
J. Lang
Discussants: Marta Kutas, Gregory A. Miller
The participants in this panel serve on the latest version of the Committee
on the Future of Psychophysiology. In this panel we plan to share with the
membership our thoughts and report on our actions. We will particularly
focus on our response to the challenge presented by some officials who suggest
that with the advent of "neuroimaging" we have reached a stage
where "traditional psychophysiology" has become effectively obsolete.
It is perhaps no surprise that the committee did not find this argument
compelling. Exactly how to deal with this view will be one subject of the
panel presentation. Other challenges and opportunities facing the field
will also be discussed.